ENH: Expose symlog scaling in plotting API (#24968)

Author: charlesdong1991

doc/source/whatsnew/v0.25.0.rst

@@ -23,7 +23,7 @@ including other versions of pandas.
 
 Other Enhancements
 ^^^^^^^^^^^^^^^^^^
-
+- :func:`DataFrame.plot` keywords ``logy``, ``logx`` and ``loglog`` can now accept the value ``'sym'`` for symlog scaling. (:issue:`24867`)
 - Added support for ISO week year format ('%G-%V-%u') when parsing datetimes using :meth: `to_datetime` (:issue:`16607`)
 - Indexing of ``DataFrame`` and ``Series`` now accepts zerodim ``np.ndarray`` (:issue:`24919`)
 - :meth:`Timestamp.replace` now supports the ``fold`` argument to disambiguate DST transition times (:issue:`25017`)

pandas/plotting/_core.py

@@ -287,8 +287,10 @@ def _maybe_right_yaxis(self, ax, axes_num):
             if not self._has_plotted_object(orig_ax):  # no data on left y
                 orig_ax.get_yaxis().set_visible(False)
 
-            if self.logy or self.loglog:
+            if self.logy is True or self.loglog is True:
                 new_ax.set_yscale('log')
+            elif self.logy == 'sym' or self.loglog == 'sym':
+                new_ax.set_yscale('symlog')
             return new_ax
 
     def _setup_subplots(self):
@@ -310,10 +312,24 @@ def _setup_subplots(self):
 
         axes = _flatten(axes)
 
-        if self.logx or self.loglog:
+        valid_log = {False, True, 'sym', None}
+        input_log = {self.logx, self.logy, self.loglog}
+        if input_log - valid_log:
+            invalid_log = next(iter((input_log - valid_log)))
+            raise ValueError(
+                "Boolean, None and 'sym' are valid options,"
+                " '{}' is given.".format(invalid_log)
+            )
+
+        if self.logx is True or self.loglog is True:
             [a.set_xscale('log') for a in axes]
-        if self.logy or self.loglog:
+        elif self.logx == 'sym' or self.loglog == 'sym':
+            [a.set_xscale('symlog') for a in axes]
+
+        if self.logy is True or self.loglog is True:
             [a.set_yscale('log') for a in axes]
+        elif self.logy == 'sym' or self.loglog == 'sym':
+            [a.set_yscale('symlog') for a in axes]
 
         self.fig = fig
         self.axes = axes
@@ -1900,12 +1916,18 @@ def _plot(data, x=None, y=None, subplots=False,
         Place legend on axis subplots
     style : list or dict
         matplotlib line style per column
-    logx : bool, default False
-        Use log scaling on x axis
-    logy : bool, default False
-        Use log scaling on y axis
-    loglog : bool, default False
-        Use log scaling on both x and y axes
+    logx : bool or 'sym', default False
+        Use log scaling or symlog scaling on x axis
+        .. versionchanged:: 0.25.0
+
+    logy : bool or 'sym' default False
+        Use log scaling or symlog scaling on y axis
+        .. versionchanged:: 0.25.0
+
+    loglog : bool or 'sym', default False
+        Use log scaling or symlog scaling on both x and y axes
+        .. versionchanged:: 0.25.0
+
     xticks : sequence
         Values to use for the xticks
     yticks : sequence

pandas/tests/plotting/test_frame.py

@@ -245,16 +245,34 @@ def test_plot_xy(self):
         # TODO add MultiIndex test
 
     @pytest.mark.slow
-    def test_logscales(self):
+    @pytest.mark.parametrize("input_log, expected_log", [
+        (True, 'log'),
+        ('sym', 'symlog')
+    ])
+    def test_logscales(self, input_log, expected_log):
         df = DataFrame({'a': np.arange(100)}, index=np.arange(100))
-        ax = df.plot(logy=True)
-        self._check_ax_scales(ax, yaxis='log')
 
-        ax = df.plot(logx=True)
-        self._check_ax_scales(ax, xaxis='log')
+        ax = df.plot(logy=input_log)
+        self._check_ax_scales(ax, yaxis=expected_log)
+        assert ax.get_yscale() == expected_log
+
+        ax = df.plot(logx=input_log)
+        self._check_ax_scales(ax, xaxis=expected_log)
+        assert ax.get_xscale() == expected_log
+
+        ax = df.plot(loglog=input_log)
+        self._check_ax_scales(ax, xaxis=expected_log, yaxis=expected_log)
+        assert ax.get_xscale() == expected_log
+        assert ax.get_yscale() == expected_log
+
+    @pytest.mark.parametrize("input_param", ["logx", "logy", "loglog"])
+    def test_invalid_logscale(self, input_param):
+        # GH: 24867
+        df = DataFrame({'a': np.arange(100)}, index=np.arange(100))
 
-        ax = df.plot(loglog=True)
-        self._check_ax_scales(ax, xaxis='log', yaxis='log')
+        msg = "Boolean, None and 'sym' are valid options, 'sm' is given."
+        with pytest.raises(ValueError, match=msg):
+            df.plot(**{input_param: "sm"})
 
     @pytest.mark.slow
     def test_xcompat(self):

pandas/tests/plotting/test_series.py

@@ -567,16 +567,21 @@ def test_df_series_secondary_legend(self):
         tm.close()
 
     @pytest.mark.slow
-    def test_secondary_logy(self):
+    @pytest.mark.parametrize("input_logy, expected_scale", [
+        (True, 'log'),
+        ('sym', 'symlog')
+    ])
+    def test_secondary_logy(self, input_logy, expected_scale):
         # GH 25545
         s1 = Series(np.random.randn(30))
         s2 = Series(np.random.randn(30))
 
-        ax1 = s1.plot(logy=True)
-        ax2 = s2.plot(secondary_y=True, logy=True)
+        # GH 24980
+        ax1 = s1.plot(logy=input_logy)
+        ax2 = s2.plot(secondary_y=True, logy=input_logy)
 
-        assert ax1.get_yscale() == 'log'
-        assert ax2.get_yscale() == 'log'
+        assert ax1.get_yscale() == expected_scale
+        assert ax2.get_yscale() == expected_scale
 
     @pytest.mark.slow
     def test_plot_fails_with_dupe_color_and_style(self):